Protracted abstinence can be defined as a state of heightened reactivity to aversive and appetitive stimuli long after acute withdrawal that conveys a vulnerability to relapse in individuals with a history of dependence. The purpose of this proposal is to establish animal and human models of protracted abstinence that can predict efficacy of medications to treat the protracted abstinence stage of the alcoholic syndrome and protect against relapse. Animal models exist for all stages of the addiction cycle and have been useful in elucidating the neurobiological basis for vulnerability to alcoholism. Preliminary studies have shown that rats can be trained to self-administer alcohol and, when made dependent, increase their consumption during acute withdrawal and long after acute withdrawal. Studies are proposed to further characterize and develop this animal model of protracted abstinence and to attempt to manipulate alcohol intake with appetitive and aversive stimuli (Specific Aim 1). Neuropharmacological agents known to alter reward dysregulation will be tested on these models (Specific Aim 1). Optimal parameters from this model will be used to identify novel potential treatments for human laboratory testing (Specific Aim 2). In parallel, a human experimental model of protracted abstinence will be developed using cue reactivity and mood induction techniques (Specific Aim 3). Primary outcome measures are affective state and urge to drink with confirmatory physiological measures. The optimal parameters of the human laboratory model will be used to evaluate potential treatments for the protracted abstinence syndrome (Specific Aim 4), drawing on the pharmacological agents identified in the animal models of Specific Aim 2. The state of protracted abstinence in alcoholism is critical to vulnerability to relapse, and the results of the present proposed studies will provide novel treatments for protracted abstinence and thus unique approaches to alcoholism treatment.